Interaction between cytochrome P450 2A6 and Catechol-O-Methyltransferase genes and their association with smoking risk in young men

Tobacco smoking is a multi-factorial behavior with both genetic and environmental determinants [1]. Genetic factors have a greater influence on smoking cessation than do environmental factors [2]. Genetic factors are responsible for 30–50% of the variance in the risk of withdrawal symptoms, 40–75% of the variation in smoking initiation, 50% of the variance in cessation success, and 70–80% of the variation in smoking maintenance [1]. Genetic risk information enhances the motivation for smoking cessation [3]. Therefore, assessment of genetic background could be a promising tool to understand smoking risk and to guide the selection of the most effective cessation treatment for an individual smoker.

Nicotine is the major psychoactive ingredient in tobacco, and it modulates dopamine activity in the midbrain, which contributes to the development and maintenance of rewarding behaviors such as smoking [4]. Smokers modulate their smoking to maintain brain nicotine levels within a certain concentration range, and factors that alter nicotine clearance affect smoking behavior [4]. Individuals who eliminate nicotine rapidly are less likely to achieve low craving scores even after smoking freely [4]. Consequently, genetic polymorphisms in both nicotine metabolism and dopamine catabolism genes influence smoking status, interact with each other to result in risk modulation, and affect smoking cessation therapies.

The cytochrome P450 (CYP) 2A6 gene, located on chromosome 19q12-q13.2, consists of nine exons. It is involved in producing a 494-amino-acid protein that oxidizes coumarin, nicotine, and tobacco-specific nitrosamines [5]. CYP2A6 is the primary human enzyme involved in nicotine metabolism [4]. CYP2A6 catalyzes the C-oxidation of nicotine to the inactive metabolite cotinine and the subsequent conversion of cotinine into trans-3′-hydroxycotinine [4]. CYP2A6 is the most studied enzyme involved in both adult and adolescent smokers [6]. The results of studies among Taiwanese individuals indicate that the variant status of CYP2A6 is different from that among other ethnic groups [7]. Therefore, we hypothesized that, for Taiwanese individuals, polymorphisms in the CYP2A6 gene that affect smoking status could be different from those in other ethnic groups.

The catechol-O-methyltransferase (COMT) gene is located on chromosome 22q11.21, has eight exons, and produces a 271-amino-acid protein that metabolizes catecholamines [8]. Low enzyme activity of the Met allele at codon 108/158 (in the rs4680 polymorphism) of the COMT gene, which encodes a key enzyme involved in the metabolic inactivation of dopamine, has been associated with nicotine dependence [9]. Another polymorphism of COMT, rs165599, has been related to the response to bupropion therapy for smoking cessation [10]. However, those associations have been inconsistent among ethnic groups [10, 11]. Thus, it seems necessary to perform a genetic study of smoking status for each ethnicity.

Smoking addiction is currently a significant social problem in Taiwan [12]. Nonetheless, very few genetic investigations of nicotine–dopamine metabolism and smoking status among Taiwanese individuals have been performed. Recently, for Taiwanese smokers in a group of methadone maintenance patients, polymorphisms of the μ-opioid receptor gene were associated with the plasma concentration of cotinine [13]. Very recently, our team reported that the interaction of the dopamine D2 receptor (DRD2) TaqIB and monoamine oxidase A (MAOA) affected smoking intensity in young men [14]. Those findings encouraged us to perform more genetic studies of Taiwanese smokers. In relation to nicotine–dopamine metabolism, smoking status has been reported to correlate better with some gene–gene interactions than with a single gene only. These interactions include COMT and MAOA [15, 16], MAOA and CYP2A6 [17], CYP2A6 and the nicotine acetylcholine receptor gene [18, 19], and CYP 2A6 and DRD2 TaqIA [20]. However, the effects of CYP2A6 and COMT together have not been explored, and we hypothesized that a single polymorphism as well as an interaction of the two genes could be involved in smoking status.

In women not using oral contraceptives, nicotine and cotinine clearance is 13 and 24% higher, respectively, than in men [6]. Sex differences exist for cravings, affect, and preference for immediate smoking after cue exposure [21]. Because Taiwanese men smoke significantly more cigarettes than Taiwanese women [12, 22], only men were invited to participate in our study. In this study, the polymorphisms CYP2A6*1A (wild-type), CYP 2A6*1B (polymorphism of faster nicotine clearance) [23], CYP2A6*4C (the most studied polymorphism of decreased nicotine metabolism) [24], COMT Val/Met (rs4680) [9], and COMT rs165599 [10] were determined. Our objectives were to investigate the effects of the interaction between the CYP 2A6 and COMT genes and their association with smoking risk in young Taiwanese men.

The distributions of CYP2A6 and COMT polymorphisms are shown in Table 1. Using the frequencies of CYP2A6*1A/*1A, CYP2A6*1A/*1B, and CYP2A6*1B/*1B as examples, the distributions of CYP2A6 genotypes agreed with Hardy–Weinberg equilibrium: P = 0.713 and 0.332 for current smokers and never smokers, respectively. The distribution of COMT rs165599 genotypes also agreed with Hardy–Weinberg equilibrium: P = 0.966 and 0.668 for current smokers and never smokers, respectively. The distribution of COMT rs4680 genotypes did not agree with Hardy–Weinberg equilibrium: P = 0.003 and 0.012 for current smokers and never smokers, respectively.

As shown in Table 2, after using never smokers as the reference group, the OR was significantly lower in the subjects carrying the genotype of CYP2A6 low activity/variant COMT rs4680 than in those possessing the genotype of CYP2A6 wild-type/variant COMT rs4680 (0.44, 95% CI 0.19–0.98, P = 0.043). The other 14 ORs were not statistically significant (P = 0.078–0.988).

The average number of cigarettes per day among the 261 current smokers was 10. There were 127 heavy smokers and 131 light smokers (data for three smokers were missing). With light smokers as the reference group, 15 ORs for heavy smoking were not statistically significant (P = 0.061–0.112, data in Additional file 2: Table S2).

For the 261 current smokers, FTND, PCDS, and CWS-21 were analyzed for 181, 227, and 210 subjects, respectively, because of missing data. Each of the eight comparisons for FTND, PCDS, and CWS-21 among the groups with different CYP2A6/COMT polymorphisms was not significantly different (P = 0.224–0.911, 0.054–0.700, and 0.075–0.836, respectively) (data in Additional file 3: Table S3).

Of the 261 current smokers, 122 subjects provided urine samples for the determination of nicotine and cotinine levels by GC/MS. The values for adjusted nicotine concentration and adjusted cotinine concentration were not normally distributed in either genotype group. The SD was too high: near the mean value or even greater than the mean value. Therefore, the Mann–Whitney U test or Kruskal–Wallis test with Bonferroni adjustment was used for comparison of the values of adjusted nicotine concentration and adjusted cotinine concentration between groups carrying different genotypes. As shown in Table 3, the adjusted urinary nicotine concentration did not differ significantly in any of the comparisons. The median adjusted urinary cotinine concentration differed between the COMT rs4680 wild-type group and the COMT rs4680 variant group (92.46 ng/μL vs. 118.24 ng/μL, P = 0.041) and between the COMT rs4680 wild-type/COMT rs165599 variant group and the COMT rs4680 variant/COMT rs165599 variant group (97.10 ng/μL vs. 122.18 ng/μL, P = 0.022). The other comparisons were not statistically significant.

All subjects in this study were young adult university students. Alcohol and drug abusers were excluded. Therefore, the effects of age, education status and alcohol and drug interactions on smoking were absent.

Our results show that the distribution of COMT rs4680 does not agree with Hardy–Weinberg equilibrium, consistent with reports that Asiatic individuals have Hardy–Weinberg disequilibrium of the COMT rs4680 polymorphism [8]. The main finding of this study is that among adult males with variant COMT rs4680, the subjects carrying the low-activity genotype of CYP2A6 have a 0.44-fold lower risk of starting smoking than those possessing the wild-type genotype of CYP2A6. In other words, the OR is 2.27-fold higher in subjects carrying the CYP2A6 wild-type/variant COMT rs4680 genotype than in those possessing the CYP2A6 low activity/variant COMT rs4680 genotype. The CYP2A6 and COMT genes were reported to be associated with smoking status. CYP2A6*4C is a whole-deletion type, decreased nicotine metabolism polymorphism [34]. In Japanese adults and young Japanese students, the frequency of the CYP2A6*4C gene was significantly higher among non-smokers than smokers [34, 35]. Among Chinese males, participants with the CYP2A6*4C genotype had a lower risk of smoking initiation and smoking persistence than those with the CYP2A6*1/CYP2A6*1 genotype [17]. For people living in southern China, reduced metabolic function of CYP2A6 in smokers appears to be associated with fewer cigarettes smoked, later initiation of smoking regularly, shorter smoking duration, and lower likelihood of smoking cessation [36]. For Caucasian individuals, the CYP2A6 slow inactivator genotype increased the risk of nicotine dependence when smoking was initiated during adolescence. However, it reduced the risk of smoking initiation, lowered cigarette consumption, and decreased the duration of smoking among adult dependent smokers [37]. The continued effect of slow metabolism on reducing cigarette consumption, throughout the smoking history of people with CYP2A6 slow inactivators, may affect tolerance and withdrawal mechanisms among these individuals [37]. Smokers with CYP2A6 slow inactivators smoke fewer cigarettes and tend to be less dependent on nicotine than smokers with normal activity alleles [38]. With respect to smoking initiation, adolescents with slower activity alleles may progress to nicotine dependence more slowly than normal metabolizers [38]. Very recently, researchers reported that CYP2A6 slow metabolism was associated with increased adolescent smoking cessation in Caucasian individuals [39]. The ORs for current smoking are reportedly higher in COMT rs4680 G/G (the high activity allele) carriers than in those possessing the COMT rs4680 variant among healthy Caucasian men of Croatian origin [40], Americans of European ancestry [41], and Japanese men [15]. On the other hand, the COMT rs4680 variation (G/A or A/A, the low-activity alleles) was associated with nicotine dependence in men and women of African-American and European-American descent [9], smoking severity among Chinese male smokers [42], heaviness of smoking in Caucasian pregnant women [43], and susceptibility to cigarette smoking among Thai males [11]. Our results demonstrate for the first time that the combination of the low-activity CYP2A6 genotype and low-activity COMT genotype is associated with the risk of starting smoking.

We also found that the adjusted urinary cotinine concentration was higher in subjects with low-activity COMT genes than in those with high-activity COMT genes. Additionally, it was higher in subjects with the COMT rs4680 variant/COMT rs165599 variant than in those with the COMT rs4680 wild-type/COMT rs165599 variant. The urinary cotinine concentration is a reliable easy-to-use marker for plasma levels of cotinine and the sum of nicotine metabolites in smokers [44]. Therefore, our findings indicate that cotinine and the sum of nicotine metabolites are metabolized more slowly in subjects carrying the low-activity genotype than in those possessing the high-activity genotype. However, such a difference did not affect FTND, PCDS, or CWS-21 scores, and it was not related to the risk of heavier smoking. The smoking intensity of the university students was lower (average cigarettes/day = 10 and average FTND score = 3.7) than that of Taiwanese adult smokers, e.g., average cigarettes/day = 24 and average FTND score = 7.1 for Chinese adults [42]. This finding may be the reason, at least in part, that the low activity genotypes of CYP2A6 and COMT were associated with high nicotine dependence scores and heavier smoking, as we reported in this study. For Taiwanese university students, the primary reason for the first contact with smoking was curiosity. Anxiety, avoiding stress, and the difficulties of smoking cessation explained continuing smoking behavior among university students [45]. The environmental factors may be more predominant than certain genetic factors (e.g., polymorphisms of CYP2A6 and COMT genes) among Taiwanese smokers in the university setting.

There are several limitations to our study. The percentage of female smokers in the Taiwanese population is low, only 4.4%, and they are difficult to recruit as participants. In our report, study participation was restricted to young male Taiwanese smokers. Although the total number of study subjects was relatively large, the numbers of individuals with some of the genotypes were too small to reach statistically significant power. Therefore, further studies with a larger sample size and that include female smokers are needed.

A single nucleotide polymorphism (rs4680) in the COMT gene and the interaction between the CYP2A6 and COMT genes affect smoking status in young Taiwanese men. Effective smoking prevention and cessation intervention programs are required to reduce smoking among university students [46]. We found that the interaction of the low-activity CYP2A6 genotype and low-activity COMT genotype is associated with the risk of starting smoking. In addition, interaction of the DRD2 TaqIB and MAOA genes also affects smoking intensity in Taiwanese young men [14]. This knowledge is useful for developing an approach to reducing smoking among Taiwanese university students. A clearer understanding of the relative roles of genetic and non-genetic factors in the initiation of smoking could have implications for the design of smoking prevention programs [47].